I can not now locate the thread in which I posted the idea that we could add some letters borrowed from Cyrillic to the Latin alphabet to have distinctive letters for the unvoiced b, g, and d sounds in Chinese.
Anyways, my recent studies of the Brahmi-derived abugidas of the East have suggested another idea to me.
Tibetan places a dot between each syllable, and it seems to me that one of the reasons Devanagari uses ligatures extensively is also to make the division between syllables unambiguous.
I have been thinking of how the division between upper- and lower- case in the Latin alphabet could be used for the same purpose.
Instead of having any need for a virama symbol, things would work like this:
Small capitals, having no ascenders or descenders, would stand for consonants plus an inherent -a vowel following.
Small capital A would stand for the "null" consonant.
Accents would be used to select other vowels - Á would be e, À would be i, Â would be o, and Ä would be u. But while those accents would be used on consonants like K, in the case of A, instead of accents, Á would be written as small-capital E, À would be written as small-capital И, Â would be written as small-capital O, and Ä would be written as small capital U.
И is used instead of I because small-capital I following a small-capital letter is used to form a dipthong, indicating a long vowel. To follow existing practice (in Devanagari, to which this would be essentially isomorphic) the accent goes on the I instead of the preceding letter.
Capital letters, which always ascend, not leaving as much room for accents, would indicate pure consonants - thus avoiding the virama. A capital letter would be used to indicate the last consonant of a syllable if there is such a consonant after the vowel.
All other consonants in the syllable, whether consonants before the consonant immediately preceding the vowel (the small capital one) or after the vowel but prior to the last consonant, would be in lower case.
Where the lower-case letter is not sufficiently distinct from the small capital, an alternate is found; i.e., lower-case s could possibly be the long s.
This way, speakers of languages such as Tibetan, Burmese, Thai, Hindi, Gujarati... could all use the same alphabet, without having to change how they spell, and they would be using glyphs with (most of) which many of them would already be familiar.
So, basically, the alphabet would be like this:
A(अ) E(ए) И(इ) O(ओ) U(उ)
AI(आ) EI(ऐ) ИI(ई) OI(औ) UI(ऊ)
K(क) X(ख) G(ग) Г(घ) ?(ङ)
Ч(च) Ջ(छ) J(ज) Ж(झ) Ն(ञ)
Δ(ट) Θ(ठ) Դ(ड) Ð(ढ) ?(ण)
T(त) Þ(थ) D(द) Д(ध) N(ऩ)
P(प) П(फ) B(ब) Б(भ) M(म)
Y(य) R(र) L(ल) Л(ळ) V(व)
Ш(श) Շ(ष) S(स) H(ह)
To get four kinds of N, though, since the Cyrillic N looks like H, I'm at a loss for what alphabets to raid. Apparently, ण has changed its form, as I expected a slanting stroke at the bottom, starting from the first vertical line, not touching the other two vertical lines (instead of looping into the second)... initially, I had some trouble finding all the Devanagari characters I was trying to match.